1. Field of the Invention
The present invention relates to a color suppression circuit and an electronic camera using it. More particularly, the invention relates to a color suppression circuit capable of suppressing a high-luminance part of a signal supplied from a CCD imager having color filters for complementary colors and used in a still camera.
2. Description of the Background Art
Referring to FIG. 3A or 3B, in a CCD having color filters for complementary colors, the color filters for four complementary colors (yellow (Ye), cyan (Cy), green (Gr), magenta (Mg)) are arranged.
FIG. 4 is a block diagram of a still camera using a CCD where the complementary color filters illustrated in FIGS. 3A and 3B are arranged. Referring to FIG. 4, an output signal of a CCD 1 is supplied to a correlation double sampling circuit (hereinafter referred to as CDS circuit) 2 to reduce noise, and thereafter converted to a 10-bit digital signal by an A/D converter 3. In the case of a single-plate camera, the signal from A/D converter 3 is supplied to a pixel interpolation circuit 4, and four complementary color signals Ye, Cy, Mg and Gr are generated at all pixel positions. The signals are thereafter supplied to a matrix (hereinafter referred to as MTX) circuit 5, and luminance signal Y1 and color difference signals CR and CB represented by the following equations (1a), (1b) and (1c) are produced.
Y1=Ye+Mg+Cy+Gr(=2r+3g+2b)xe2x80x83xe2x80x83(1a)
CR=(Ye+Mg)xe2x88x92(Cy+Gr)(=2rxe2x88x92g)xe2x80x83xe2x80x83(1b)
CB=(Mg+Cy)xe2x88x92(Gr+Ye)(=2bxe2x88x92g)xe2x80x83xe2x80x83(1c)
The three signals are then supplied to an MTX circuit 6 and converted to r, g and b by the following equations (2a), (2b) and (2c).
r=0.1(Y1+4CRxe2x88x92CB)xe2x80x83xe2x80x83(2a)
g=0.1(2Y1xe2x88x922CRxe2x88x922CB)xe2x80x83xe2x80x83(2b)
b=0.1(Y1xe2x88x924CR+CB)xe2x80x83xe2x80x83(2c)
The converted r, g and b signals are supplied to a white balance circuit 7. R gain and B gain are supplied to white balance circuit 7 from a white balance control circuit 11, and white balance circuit 7 multiplies the R gain and the r signal, and multiplies the B gain and the b signal to carry out white balance correction, and outputs RGB signals represented by equations (3a), (3b) and (3c) below.
R=R gainxc3x97rxe2x80x83xe2x80x83(3a)
G=gxe2x80x83xe2x80x83(3b)
B=B gainxc3x97bxe2x80x83xe2x80x83(3c)
Overflow-clipping is applied to the RGB signals to produce full scale 10 bits by white balance circuit 7, and the resultant RGB signals are supplied to a gamma correction circuit 8 where gradation correction is made by 0.45th power to generate 8-bit data.
Rxe2x80x2=255xc3x97(R/1023){circumflex over ( )}0.45xe2x80x83xe2x80x83(4a)
Gxe2x80x2=255xc3x97(G/1023){circumflex over ( )}0.45xe2x80x83xe2x80x83(4b)
Bxe2x80x2=255xc3x97(B/1023){circumflex over ( )}0.45xe2x80x83xe2x80x83(4c)
White balance control circuit 11 calculates the average in the picture plane, sumR, sumG and sumB for each data of R, G and B and feeds back the R gain and G gain to a multiplier circuit of white balance circuit 7 such that sumR=sumG and sumB=sumG are satisfied. The primary color signals Rxe2x80x2, Gxe2x80x2 and Bxe2x80x2 to which gamma correction has been applied by gamma correction circuit 8 are supplied to a matrix circuit 9 and converted to a luminance signal Y and color difference signals U and V according to the following equations (5a), (5b) and (5c).
Y=0.299Rxe2x80x2+0.587Gxe2x80x2+0.114Bxe2x80x83xe2x80x83(5a)
U=xe2x88x920.1684(Rxe2x80x2xe2x88x92Gxe2x80x2)+0.5(Bxe2x80x2xe2x88x92Gxe2x80x2)xe2x80x83xe2x80x83(5b)
V=0.5(Rxe2x80x2xe2x88x92Gxe2x80x2)xe2x88x920.0813(Bxe2x80x2xe2x88x92Gxe2x80x2)xe2x80x83xe2x80x83(5c)
For a high-luminance subject, all four complementary colors of the output from CCD 1 are saturated. If 10-bit A/D converter 3 is used, Ye=Cy=Mg=Gr=1023 is established. Each signal amount is then determined as below using equations (1) to (5).
CR=CB=0xe2x80x83xe2x80x83(6)
Y1=4092, r=b=409, g=818xe2x80x83xe2x80x83(7)
Suppose that the white balance correction values given by white balance control circuit 11 are obtained as below.
R gain=2, B gain=2xe2x80x83xe2x80x83(8)
Then the following result is obtained.
R=G=B=818
Rxe2x80x2=Gxe2x80x2=Bxe2x80x2=231
Y=231
U=V=0
Accordingly, the corresponding portion is achromatic color.
In most cases, the spectral characteristics of the complementary color filters are determined as shown in FIG. 5 such that CR=CB=0 is satisfied as shown in the equation (6) when a white subject is imaged with illumination by ordinary white light (color temperature 5500K). Following the equation (2), the expressions below are satisfied.
r=0.1Y1, g=0.2Y1, b=0.1Y1xe2x80x83xe2x80x83(9)
Consequently, the multipliers for the white balance correction are obtained as those values of the equation (8).
Considering above, when a white subject is imaged with illumination at color temperature 5500K, a non-saturated portion and a saturated portion are both reproduced as achromatic color if R gain=2 and B gain=2 are applied.
In the case of illumination by an incandescence lamp (color temperature 2750K) for a white non-saturated subject, white balance correction values are obtained as R gain=1 and B gain=4 since r=g=818 and b=205 are applied. However, for the portion where all four complementary colors are saturated, equations (6) and (7) are satisfied as in the case of color temperature 5500K. Therefore, following expressions are established after the white balance correction.
R=1xc3x97409=409, G=818, B=4xc3x97409=1023(due to overflow clipping)
Accordingly, when the white balance correction is precisely applied to the non-saturated portion of the incandescence lamp, the saturated portion is colored to bluish green. On the contrary, if white balance correction values are obtained as R gain=4 and B gain=1 as in the case of the shade in fine whether (color temperature 11000K), white balance output in the saturated portion is R=1023, G=818, and B=409, and the portion is colored to yellowish red.
In order to solve this problem, a color suppression circuit 10 and a luminance.color gain conversion circuit 12 are conventionally provided for color suppression of a high-luminance part as shown in FIG. 4. Luminance.color gain conversion circuit 12 sets a multiplier of UV (or modulation color signal) or the gain of an amplifier to color suppression circuit 10 based on luminance signal Y1. However, this operation is not enough to appropriately carry out suppression in a wide range of color temperature. Specifically, if a sufficient suppression effect is to be ensured for color temperatures 2750K and 11000K, suppression is excessively done for 5500K to deteriorate the dynamic range.
One object of the present invention is to provide a color suppression circuit to obtain an excellent color suppression characteristic for a high-luminance part relative to variation in color temperature and provide an electronic camera employing the color-suppression circuit.
Briefly, the present invention is implemented as a color suppression circuit capable of suppressing a high-luminance part of a signal supplied from an imager where color filters are arranged for respective pixels. The signal supplied from the imager is processed to extract a luminance signal and information on white balance from a signal extraction circuit. Based on the extracted luminance signal and white balance information, the color gains of the high-luminance part of the signal from the imager is controlled. According to the invention, a superior color suppression characteristic for the high-luminance part can thus be obtained relative to change in color temperature, by controlling the color gains of the high-luminance part.
According to a preferred embodiment of the invention, a signal extraction circuit includes an A/D conversion circuit converting the signal from the imager to a digital signal, a matrix circuit generating a luminance signal and color signals of RGB from the converted digital signal, a white balance circuit applying white balance adjustment to the signals of RGB from the matrix circuit, and a white balance control circuit extracting information on white balance from color information of RGB of a white balance circuit output.
According to a more preferred embodiment of the invention, a suppression circuit is formed of a gain conversion circuit outputting a gain conversion signal for decreasing the color gains of the high-luminance part based on the luminance signal from the matrix circuit and the white balance information from the white balance control circuit, and a color suppression circuit applying color suppression to a high-luminance part of a color signal supplied from the white balance circuit, based on the supplied gain conversion signal.
According to a preferred embodiment of the invention, the signal suppression circuit is contained in an electronic camera.
The foregoing and other objects, features, aspects and advantages of the present invention will become more apparent from the following detailed description of the present invention when taken in conjunction with the accompanying drawings.